Analysis And Study For Collecting Methods Of The Highway Traffic Basic Information Space Data

At present, there are three methods in collecting the highway transportation basic space data: collecting it by relief map, RS and GPS. In our country, the relief map is the result of reconnaissance on the spot which has the characteristics of a higher in geometry precision and stabilization in the data of physiognomy and drainage areas on average. At the same time, the position of all classes of highway is relatively stable. Therefore, collecting data by relief map is the means. But some of the highways can't be included after the article is born, that's to say, this means has a certain limitation. The data isn't complete and can't reflect the present conditions. Compared to the relief map on the paper, remote sensing pictures are short in circulation (nowadays LandSat circles the earth once every 16 days. that is to say, a new picture information can be got every 16 days. we can dispel out highway space data by sensing pictures as the completion of relief map data, in order to supply the recent changes and keep the completion of the highway space data. The purpose is to reflect the present highway conditions. Using GPS to collect data is a precise and practical way (to the actual route) but the way coat lots of labour and money. Because the changes of all levels aren't very much, It's unnecessary to coat so much. The major work of the thesis is to study and analyze the means of collecting highway transportation basic information. Put forward the way that we can take advantage of the newest method-using relief map and LandSat EMT to correct and renew it. Because the relief map which includes highway, railway, contour, rivers and lakes and so on is generally very complicated and it is because that the time, position, quality of printing and scanning can influence its precision it results in the high possibility of mistakes. So it can not put to use. As a result, we take the step of half-motion tracking by scanning the relief into our computer through scanner firstly. And then deal with the original pictures. Say, smooth away the noise and its rough factor. Because there is no mark on the relief map, while the position is vectored, we only can take half-motion track this causes lots of different pictures with different vectors separately. They aren't connected with each other in space. So we createa shape document with 1:50000 and then projects and switches it under ArcView. We correct every map and half-motion tracks it, at last we get maps with vector pictures. The pictures are space piled analyzed added to vectorgraphs. The pictures are corrected in space and pick up the information through vectorization software. The relief map is made of four-colour overprinted. Because the colour pictures differ greatly in colour, the half-motion track hasn't a good continuous characteristic while using vectorization software, meanwhile the speed is not fast. When the half-motion track is carried on, the colourful pictures are switched into two-twain pictures. But the later pictures contain all kinds of information including highways, waters and contour, so it causes difficulties in recognition of the objects. On the map with colours, we can recognize a blue river, a black highway, a brown contour. However, on the changed mapâ…¡these objects are marked with dark black. We can't tell the differences according to the colour again. Only according to the shape can we tell them. But the three factors are all line-shaped ones, so they are easily confused. The case lows its recognition and therefore easily causes fail to recognize highways. It enlarges the difficulty in digitalizing. Her we put forward a delamination-pick-up method firstly and deal with different levels of pictures separately. In this way can we adopt different data by different colour. The information types contain brown contour, blue waters and lakes, black transportation lines.